Guiding a star on spectroscope slit

Main forum about Astroart
Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Guiding a star on spectroscope slit

Post by Forum_2014 »

Hello Fabio, concerning tools in support of spectroscopy, I would like to kindly ask you for a specific guiding feature. Would be nice if AA could implement such a thing in future.

Many amateur, like me, use a reflective slit placed on telescope focal plane that allows to check (with a dedicated camera) the field of view and center the target star exactly on the slit. Up to mag 11-12 (but depending on setup) the star is still visible when centered in the slit. More precisely you see the star image halo that appears divided in two parts, being cut in the center by slit. This is an example with a very bright guide star:
slit_guide.jpg
I tried to guide a so shaped image with usual parameters but the star become not stable swinging above and below the slit (placed horizontally). This greatly reduces the efficiency because the light does not enter all time into spectroscope. I tried to change parameters (dead zone and speed) but I could not eliminate the problem.

I read that some spectroscopy softwares use the star's halo, not a centroid calculation, to optimize the guide. For example by dividing in four sectors a small area centered on the guide point and sending corrections to mount in order to mantain a balanced intensity halo in each sector.

Another essential prerequisite is that you can enter manually the coordinates of the guiding point (better if subpixel). A point that we have verified lying along the slit projected on guide camera sensor.


There would be a few features very useful for faint target that disappear completely when in slit but there are some other stars in field of view (used for guiding). Essential, these functions should be also accessible while CCD control is guiding.

- shift the guide point along X or Y (if the target come into view emerging from the slit - e.g. for fine centering of a star or asteroids/comets with proper motion tracking)

- manual set of view thresholds (if the guide star is bright we'll never see the faint target with an auto-stretching leveling)

- change of guide camera exposure time (sometime, view thresholds setting may not be sufficient to see a very faint target; we have to increase exp time for a best centering on slit and then switch back to a short exposure for optimal guiding on a field star).

I think currently there are no integrated software packages with these capability that, in my opinion, are sufficient to cover any situation concerning guide on reflective slit (Ken or other, can you confirm?). I understand that we're talking about a lot of work!

Thank you in advance for your consideration. I would be avalilable for any support if you need.
Paolo

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Paolo,
+1 from me!!
(The only other package I'm aware off which does something similar (i.e. guide on overflow slit light) is SPCAudAce.)

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

It seems to me that it can be added easily and quickly.
If you can, send me by email some typical FITS images. In your image it seems that the guide star is very bright, what is the worst case? A FITS image would be useful.

1) Do you mean the subframe coordinates of the CCD?
2) This could use the same setting of "Visualization of focus frame" in the Options1 tab.
3) This works with the CCD Interface 5.21.

Thanks,
Fabio.

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Wow, that would be great!

I see your answer too late yesterday (I was observing), otherwise I would have prepared all data you need. Today is cloudy, the first chance of good weather I'll take a set of frames in fits format with different magnitude stars on slit, from bright to faint (I'll send to your e-mail).

1) Yes but I don't know how AA handles coordinate in the guide window (absolute, relative?). As said we have two circumstances I describe in more detail.

- Guide on slit (usual mode), where we select a little subframe around the star. We can check if, on average, the star is well centered in the slit according to its shape. For example if we note that the star tends to be up-shifted it means that Y coordinate entered manually was not very precise (or if we are observing an object with slow proper motion) and it would be very useful to be able to bring back the target to the correct point (in slit) without stopping autoguide. Some spectroscopist place the slit axis horizontally, other vertically. I think guide window might include 4 little buttons, +Y, -Y, +X, -X.

- Guide not on slit on a field star (target too low bright for autoguiding). We select a large subframe which includes both the guide star and the target approximately on slit (would be optimal if the guide window was not very small) then, with a relatively long exposure time and an appropriate level stretching, we could precisely center on slit the target by buttons. I was thinking that arises the problem of having two stars in the guide window with the need to guiding on one of the two (specified by us but with a good chance the bright one)...

2) It would be perfect!

3) Very well (sorry, I did not know...)

See you soon, thank you very much Fabio!
Paolo

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Since the release of PHD2 there have been some interesting discussion on slit guiding over on the astronomical spectroscopy group:
http://groups.yahoo.com/neo/groups/astr ... opics/9016

To paraphrase some of the comments:

The freeze frame feature sounds a nice idea. getting the star exactly
> on the slit PHD can be frustrating when guiding on field stars (when
> guiding on the slit I use a workround which allows the guide position
> to be fixed by clicking on the illuminated slit before starting
> guiding which hopefully will still work in PHD2)
>
>
> Perhaps even better would be the ability to couple a "guide where I
> click, not where the star is" feature with a "nudge" feature (at sub
> pixel steps) for the guide position. This would allow the star to be
> moved exactly onto the slit under guider control and held there.

Could be good for spectroscopy perhaps if someone (sadly not me) has the programming skills to add important features like "guide where I click, not where the star is" to locate the guide point on the slit and "treat a split star image as a single star" to stop the tendency to swing from side to side on the slit.

It has it's good and bad points compared to CCDSoft.
It is much more difficult to accurately position the star on the slit with PHD. I like to position the star on the slit accurately then select a guide star and start guiding on that star. This means not using continuous imaging but taking single exposures and freezing it when the star is accurately on the slit. I couldn't find an easy way to do this with PHD2. I can use single exposures with CCDSoft.
However, when I started guiding PHD uses the entire frame so I can see the slit and the guide star at the same time. This allows constant checking of the position of the target star. CCDSoft uses a small subframe around the guide star of ~20 pixels square. This is faster at downloading but means you don't have an image of the slit unless I'm guiding on the slit and target star. I only do this for bright star as dim ones dissappear on the slit. I also use CCDSoft to take the spectra so only need 1 programme running.

The end result is that both types of software work with advantages to both. CCDSoft is old software now and isn't being updated. If PHD2 can be modified to guide on the dumbell shape of a star on a slit it will have an advantage but until then I think I don't have any reason to swap.

The freeze frame feature sounds a nice idea. getting the star exactly on the slit PHD can be frustrating when guiding on field stars (when guiding on the slit I use a workround which allows the guide position to be fixed by clicking on the illuminated slit before starting guiding which hopefully will still work in PHD2)


Perhaps even better would be the ability to couple a "guide where I click, not where the star is" feature with a "nudge" feature (at sub pixel steps) for the guide position. This would allow the star to be moved exactly onto the slit under guider control and held there. Add these and a guiding algorithm which can take care of the split star image to PHD and it would be a winner.


I am currently using CCDOPS for guiding. It came free with the ST-i
camera I am using for guiding. It has the "guide where I click" and it
has a "nudge" feature although I do not know if it is at the sub pixel
level. I haven't tried the latter feature, yet, and so I don't know how
well it works.

I hope these comments give some insight into the challenges and the possible solutions to slit guiding.

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Wow Ken, really heavy climatic conditions! Also for CCDs...

Coming back to topic...

Fabio, I would like to clarify about initial operations with guide on slit, perhaps not very clear considering actual AA guiding procedure (that use a subframe of CCD sensor). I ask you any tips to find the best way to operate.

A brief preamble. I use the same X/Y coordinate (absolute) on the CCD sensor since last time I reassembled Lodestar camera on spectrograph (about six month ago). I mean that once you find the right values, you do not have to waste time looking for them again, the star can be directed precisely towards the slit using coordinates (manual entry or stored values recall). This is true as long as you do not change the position of the guide camera. In particular conditions (very seldom) I experienced a slight shift (1 pixel or less) due to mechanical flexures.

Having said this, operations to do, step by step, could be:

1) illuminate the slit (with calibration lamp, flatbox, skylight,...)
2) take a light frame (we obtain an image like the one shown in my previous post)
3) identify with mouse cursor a point along the slit (horizontally centered)
4) open the optical path to sky and take a light frame (star field with target inside)
5) select the target and start the guiding procedure (target will move toward this guide point, remaining in position during the exposure)
6) observe the stellar shape during guide (with appropriate view thresholds and exposure) and precisely center on slit with +Y -Y buttons, if possible with subpixel step
7) read current X/Y coordinates of guide point, information to use for any subsequent observations

Note that point 5 does not consider the pratical way for switching to the window guide. Entire size window? Open a subframe of preset size? Effectively does not make sense to manually select a subframe. I think you have the best idea about it...

Once you find precise X/Y coordinates, for next targets you can skip all described steps apart points 4-5. Probably this is true also for next observing sessions if the guide camera is not removed. You only need to select the target in the frame and start guide (center and mantain) function.

Please, let me know if something is missing or not right.
Paolo

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

OK, but top priority is the algorithm to guide on the splitter star. Using/setting a fixed point as guide reference is easy, it's a feature yet of the Focus Window (please confirm):

to center the guide star to a fixed coordinate on the CCD, you may use the Focus Window, watching the "X,Y" parameters on the status bar, which do not change with the subframe position (they are absolute coordinates). Does it helps?

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Sounds good to me!
I use Al's Reticule to show the slit gap position, height and a "centre" circle.
(In the files area of the astronomical spectroscopy group)
http://groups.yahoo.com/neo/groups/astr ... copy/files

Once set, this defines the slit position every time....(well almost every time!)
The issue/ problem I have is calibrating, then nudging the target star onto the slit gap and commencing guiding without loosing the star from the gap.
As mentioned it would be great to be able to set the target star to "goto nominated (X-Y) position and guide"

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

I would like to thanks Fabio for his exceptional promptness! I tested with success the prototype he sent to me, now implemented in the last CCD Interface (release 5.30):

http://www.msb-astroart.com/

Here you find a brief tutorial on basic operations:

http://quasar.teoth.it/html/spectra/AA_slit_guide.pdf

Thank you on behalf of the amateur spectroscopy community!

Paolo

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Hi Paolo and Fabio
Another thanks from me for the new slit provision . If I`ve got the tutorial correct then the steps are as follows-
1 Use a lamp to lighten the FOV through the guide ccd to show the slit and note x/y position lets say 160,200.
2 get ref star on green crosshair and slew centre of crosshair to position 160,200 (slit would not be visible now)
3 should show slight `figure of eight` star image indicating star smack on slit
start guiding
This sound ok?
Regards
Steve

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Hello Steve, I apologize for delay...

Yes, your procedure seems to me correct. Just some advice... The best way to illuminate slit is a flatbox in front of telescope (or twilight). On my Lhires III I have to be very careful with internal neon lamp because must be properly oriented otherwise the slit does not look good or appear slightly shifted (!). You should avoid a single concentrated light source (a lamp) in front of telescope aperture.

In addition to check a `figure of eight` star image (better visibile if you have proper star/slit size ratio) or the maximum brightness attenuation (because the star is into the slit), you can verify the maximum light throughput into spectroscope: using a bright star near target you may take some spectra at various Y guide sensor positions (if you have slit axis horizontally) with 1 pixel step, taking note and using the coordinates corresponding to the best spectrum SNR.

What spectrograph are you using?

Clear skies!
Paolo

Forum_2014
Posts: 253
Joined: 03 Dec 2018, 22:33

Re: Guiding a star on spectroscope slit

Post by Forum_2014 »

Hi Paolo
Sorry for delay replying!. I have an Alpy 600 with guide module (23micron slit) . This is a typical shot through my Lodestar guide ccd of the SN in m82,you can just make out the horizontal slit with the SN placed on it, shot with my ccd software & PHD.
Regards
Steve

Post Reply